Ricardo Amper, CEO of Incode, talked about the company’s approach to detecting deepfakes and preventing identity fraud, covering the company history, product architecture, real-world deployments, and emerging agent identity needs. He described Incode’s evolution from mobile AI to identity verification and its customer base across banks, telcos, marketplaces, and elections services. He framed deepfakes as an ecosystem-level threat amplified by generative AI and low-cost attack vectors, and presented a two-part defense: DeepSight, an on-device detection SDK plus transformer models that inspect phone signals and media for manipulated content, and a zero-knowledge fraud network (trust graph) that links distributed fingerprints of legitimate users without sharing PII.
He also addressed verification for autonomous agents: Incode has launched an agent identity token, plans a zero-knowledge token for privacy-preserving accountability, and intends an October rollout with permissioned verification goals.
